ironic/install-guide/source/setup-drivers.rst
Ruby Loo 168d48eb76 Update guide for PXE multi-architecture setup
This updates documentation for PXE multi-architecture setup in
the install guide. It clarifies the values for the two *_arch
config options and combines the two examples into one.

Change-Id: I10684e16dd6efcc6d17fb4a825e55286c16a20be
Related-Bug: #1582964
2016-10-13 22:57:00 -04:00

614 lines
22 KiB
ReStructuredText

Setup the drivers for the Bare Metal service
============================================
PXE setup
---------
If you will be using PXE, it needs to be set up on the Bare Metal service
node(s) where ``ironic-conductor`` is running.
#. Make sure the tftp root directory exist and can be written to by the
user the ``ironic-conductor`` is running as. For example::
sudo mkdir -p /tftpboot
sudo chown -R ironic /tftpboot
#. Install tftp server and the syslinux package with the PXE boot images::
Ubuntu: (Up to and including 14.04)
sudo apt-get install xinetd tftpd-hpa syslinux-common syslinux
Ubuntu: (14.10 and after)
sudo apt-get install xinetd tftpd-hpa syslinux-common pxelinux
Fedora 21/RHEL7/CentOS7:
sudo yum install tftp-server syslinux-tftpboot xinetd
Fedora 22 or higher:
sudo dnf install tftp-server syslinux-tftpboot xinetd
#. Using xinetd to provide a tftp server setup to serve ``/tftpboot``.
Create or edit ``/etc/xinetd.d/tftp`` as below::
service tftp
{
protocol = udp
port = 69
socket_type = dgram
wait = yes
user = root
server = /usr/sbin/in.tftpd
server_args = -v -v -v -v -v --map-file /tftpboot/map-file /tftpboot
disable = no
# This is a workaround for Fedora, where TFTP will listen only on
# IPv6 endpoint, if IPv4 flag is not used.
flags = IPv4
}
and restart xinetd service::
Ubuntu:
sudo service xinetd restart
Fedora:
sudo systemctl restart xinetd
#. Copy the PXE image to ``/tftpboot``. The PXE image might be found at [1]_::
Ubuntu (Up to and including 14.04):
sudo cp /usr/lib/syslinux/pxelinux.0 /tftpboot
Ubuntu (14.10 and after):
sudo cp /usr/lib/PXELINUX/pxelinux.0 /tftpboot
#. If whole disk images need to be deployed via PXE-netboot, copy the
chain.c32 image to ``/tftpboot`` to support it. The chain.c32 image
might be found at::
Ubuntu (Up to and including 14.04):
sudo cp /usr/lib/syslinux/chain.c32 /tftpboot
Ubuntu (14.10 and after):
sudo cp /usr/lib/syslinux/modules/bios/chain.c32 /tftpboot
Fedora/RHEL7/CentOS7:
sudo cp /boot/extlinux/chain.c32 /tftpboot
#. If the version of syslinux is **greater than** 4 we also need to make sure
that we copy the library modules into the ``/tftpboot`` directory [2]_
[1]_::
Ubuntu:
sudo cp /usr/lib/syslinux/modules/*/ldlinux.* /tftpboot
#. Create a map file in the tftp boot directory (``/tftpboot``)::
echo 're ^(/tftpboot/) /tftpboot/\2' > /tftpboot/map-file
echo 're ^/tftpboot/ /tftpboot/' >> /tftpboot/map-file
echo 're ^(^/) /tftpboot/\1' >> /tftpboot/map-file
echo 're ^([^/]) /tftpboot/\1' >> /tftpboot/map-file
.. [1] On **Fedora/RHEL** the ``syslinux-tftpboot`` package already install
the library modules and PXE image at ``/tftpboot``. If the TFTP server
is configured to listen to a different directory you should copy the
contents of ``/tftpboot`` to the configured directory
.. [2] http://www.syslinux.org/wiki/index.php/Library_modules
PXE UEFI setup
--------------
If you want to deploy on a UEFI supported bare metal, perform these additional
steps on the ironic conductor node to configure the PXE UEFI environment.
#. Install Grub2 and shim packages::
Ubuntu: (14.04LTS and later)
sudo apt-get install grub-efi-amd64-signed shim-signed
Fedora 21/RHEL7/CentOS7:
sudo yum install grub2-efi shim
Fedora 22 or higher:
sudo dnf install grub2-efi shim
#. Copy grub and shim boot loader images to ``/tftpboot`` directory::
Ubuntu: (14.04LTS and later)
sudo cp /usr/lib/shim/shim.efi.signed /tftpboot/bootx64.efi
sudo cp /usr/lib/grub/x86_64-efi-signed/grubnetx64.efi.signed \
/tftpboot/grubx64.efi
Fedora: (21 and later)
sudo cp /boot/efi/EFI/fedora/shim.efi /tftpboot/bootx64.efi
sudo cp /boot/efi/EFI/fedora/grubx64.efi /tftpboot/grubx64.efi
CentOS: (7 and later)
sudo cp /boot/efi/EFI/centos/shim.efi /tftpboot/bootx64.efi
sudo cp /boot/efi/EFI/centos/grubx64.efi /tftpboot/grubx64.efi
#. Create master grub.cfg::
Ubuntu: Create grub.cfg under ``/tftpboot/grub`` directory.
GRUB_DIR=/tftpboot/grub
Fedora: Create grub.cfg under ``/tftpboot/EFI/fedora`` directory.
GRUB_DIR=/tftpboot/EFI/fedora
CentOS: Create grub.cfg under ``/tftpboot/EFI/centos`` directory.
GRUB_DIR=/tftpboot/EFI/centos
Create directory GRUB_DIR
sudo mkdir -p $GRUB_DIR
This file is used to redirect grub to baremetal node specific config file.
It redirects it to specific grub config file based on DHCP IP assigned to
baremetal node.
.. literalinclude:: ../../ironic/drivers/modules/master_grub_cfg.txt
Change the permission of grub.cfg::
sudo chmod 644 $GRUB_DIR/grub.cfg
#. Update the bare metal node with ``boot_mode`` capability in node's properties
field::
ironic node-update <node-uuid> add properties/capabilities='boot_mode:uefi'
#. Make sure that bare metal node is configured to boot in UEFI boot mode and
boot device is set to network/pxe.
NOTE: ``pxe_ilo`` driver supports automatic setting of UEFI boot mode and
boot device on the bare metal node. So this step is not required for
``pxe_ilo`` driver.
.. note::
For more information on configuring boot modes, see boot_mode_support_.
Elilo: an alternative to Grub2
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Elilo is a UEFI bootloader. It is an alternative to Grub2, although it
isn't recommended since it is not being supported.
#. Download and untar the elilo bootloader version >= 3.16 from
http://sourceforge.net/projects/elilo/::
sudo tar zxvf elilo-3.16-all.tar.gz
#. Copy the elilo boot loader image to ``/tftpboot`` directory::
sudo cp ./elilo-3.16-x86_64.efi /tftpboot/elilo.efi
#. Update bootfile and template file configuration parameters for UEFI
PXE boot in the Bare Metal Service's configuration file
(/etc/ironic/ironic.conf)::
[pxe]
# Bootfile DHCP parameter for UEFI boot mode. (string value)
uefi_pxe_bootfile_name=elilo.efi
# Template file for PXE configuration for UEFI boot loader.
# (string value)
uefi_pxe_config_template=$pybasedir/drivers/modules/elilo_efi_pxe_config.template
iPXE setup
----------
An alternative to PXE boot, iPXE was introduced in the Juno release
(2014.2.0) of Bare Metal service.
If you will be using iPXE to boot instead of PXE, iPXE needs to be set up
on the Bare Metal service node(s) where ``ironic-conductor`` is running.
#. Make sure these directories exist and can be written to by the user
the ``ironic-conductor`` is running as. For example::
sudo mkdir -p /tftpboot
sudo mkdir -p /httpboot
sudo chown -R ironic /tftpboot
sudo chown -R ironic /httpboot
#. Create a map file in the tftp boot directory (``/tftpboot``)::
echo 'r ^([^/]) /tftpboot/\1' > /tftpboot/map-file
echo 'r ^(/tftpboot/) /tftpboot/\2' >> /tftpboot/map-file
.. _HTTP server:
#. Set up TFTP and HTTP servers.
These servers should be running and configured to use the local
/tftpboot and /httpboot directories respectively, as their root
directories. (Setting up these servers is outside the scope of this
install guide.)
These root directories need to be mounted locally to the
``ironic-conductor`` services, so that the services can access them.
The Bare Metal service's configuration file (/etc/ironic/ironic.conf)
should be edited accordingly to specify the TFTP and HTTP root
directories and server addresses. For example:
.. code-block:: ini
[pxe]
# Ironic compute node's tftp root path. (string value)
tftp_root=/tftpboot
# IP address of Ironic compute node's tftp server. (string
# value)
tftp_server=192.168.0.2
[deploy]
# Ironic compute node's http root path. (string value)
http_root=/httpboot
# Ironic compute node's HTTP server URL. Example:
# http://192.1.2.3:8080 (string value)
http_url=http://192.168.0.2:8080
#. Install the iPXE package with the boot images::
Ubuntu:
apt-get install ipxe
Fedora 21/RHEL7/CentOS7:
yum install ipxe-bootimgs
Fedora 22 or higher:
dnf install ipxe-bootimgs
#. Copy the iPXE boot image (``undionly.kpxe`` for **BIOS** and
``ipxe.efi`` for **UEFI**) to ``/tftpboot``. The binary might
be found at::
Ubuntu:
cp /usr/lib/ipxe/{undionly.kpxe,ipxe.efi} /tftpboot
Fedora/RHEL7/CentOS7:
cp /usr/share/ipxe/{undionly.kpxe,ipxe.efi} /tftpboot
.. note::
If the packaged version of the iPXE boot image doesn't work, you can
download a prebuilt one from http://boot.ipxe.org or build one image
from source, see http://ipxe.org/download for more information.
#. Enable/Configure iPXE in the Bare Metal Service's configuration file
(/etc/ironic/ironic.conf):
.. code-block:: ini
[pxe]
# Enable iPXE boot. (boolean value)
ipxe_enabled=True
# Neutron bootfile DHCP parameter. (string value)
pxe_bootfile_name=undionly.kpxe
# Bootfile DHCP parameter for UEFI boot mode. (string value)
uefi_pxe_bootfile_name=ipxe.efi
# Template file for PXE configuration. (string value)
pxe_config_template=$pybasedir/drivers/modules/ipxe_config.template
# Template file for PXE configuration for UEFI boot loader.
# (string value)
uefi_pxe_config_template=$pybasedir/drivers/modules/ipxe_config.template
#. It is possible to configure the Bare Metal service in such a way
that nodes will boot into the deploy image directly from Object Storage.
Doing this avoids having to cache the images on the ironic-conductor
host and serving them via the ironic-conductor's `HTTP server`_.
This can be done if:
#. the Image Service is used for image storage;
#. the images in the Image Service are internally stored in
Object Storage;
#. the Object Storage supports generating temporary URLs
for accessing objects stored in it.
Both the OpenStack Swift and RADOS Gateway provide support for this.
* See `Ceph Object Gateway support`_ on how to configure
the Bare Metal Service with RADOS Gateway as the Object Storage.
Configure this by setting the ``[pxe]/ipxe_use_swift`` configuration
option to ``True`` as follows:
.. code-block:: ini
[pxe]
# Download deploy images directly from swift using temporary
# URLs. If set to false (default), images are downloaded to
# the ironic-conductor node and served over its local HTTP
# server. Applicable only when 'ipxe_enabled' option is set to
# true. (boolean value)
ipxe_use_swift=True
Although the `HTTP server`_ still has to be deployed and configured
(as it will serve iPXE boot script and boot configuration files for nodes),
such configuration will shift some load from ironic-conductor hosts
to the Object Storage service which can be scaled horizontally.
Note that when SSL is enabled on the Object Storage service
you have to ensure that iPXE firmware on the nodes can indeed
boot from generated temporary URLs that use HTTPS protocol.
.. _Ceph Object Gateway support: http://docs.openstack.org/developer/ironic/deploy/radosgw.html
#. Restart the ``ironic-conductor`` process::
Fedora/RHEL7/CentOS7:
sudo systemctl restart openstack-ironic-conductor
Ubuntu:
sudo service ironic-conductor restart
PXE multi-architecture setup
----------------------------
It is possible to deploy servers of different architecture by one conductor.
To use this feature, architecture-specific boot and template files must
be configured using the configuration options
``[pxe]pxe_bootfile_name_by_arch`` and ``[pxe]pxe_config_template_by_arch``
respectively, in the Bare Metal service's configuration file
(/etc/ironic/ironic.conf).
These two options are dictionary values; the key is the architecture and the
value is the boot (or config template) file. A node's ``cpu_arch`` property is
used as the key to get the appropriate boot file and template file. If the
node's ``cpu_arch`` is not in the dictionary, the configuration options (in
[pxe] group) ``pxe_bootfile_name``, ``pxe_config_template``,
``uefi_pxe_bootfile_name`` and ``uefi_pxe_config_template`` will be used
instead.
In the following example, since 'x86' and 'x86_64' keys are not in the
``pxe_bootfile_name_by_arch`` or ``pxe_config_template_by_arch`` options, x86
and x86_64 nodes will be deployed by 'pxelinux.0' or 'bootx64.efi', depending
on the node's ``boot_mode`` capability ('bios' or 'uefi'). However, aarch64
nodes will be deployed by 'grubaa64.efi', and ppc64 nodes by 'bootppc64'::
[pxe]
# Bootfile DHCP parameter. (string value)
pxe_bootfile_name=pxelinux.0
# On ironic-conductor node, template file for PXE
# configuration. (string value)
pxe_config_template = $pybasedir/drivers/modules/pxe_config.template
# Bootfile DHCP parameter for UEFI boot mode. (string value)
uefi_pxe_bootfile_name=bootx64.efi
# On ironic-conductor node, template file for PXE
# configuration for UEFI boot loader. (string value)
uefi_pxe_config_template=$pybasedir/drivers/modules/pxe_grub_config.template
# Bootfile DHCP parameter per node architecture. (dict value)
pxe_bootfile_name_by_arch=aarch64:grubaa64.efi,ppc64:bootppc64
# On ironic-conductor node, template file for PXE
# configuration per node architecture. For example:
# aarch64:/opt/share/grubaa64_pxe_config.template (dict value)
pxe_config_template_by_arch=aarch64:pxe_grubaa64_config.template,ppc64:pxe_ppc64_config.template
Networking service configuration
--------------------------------
DHCP requests from iPXE need to have a DHCP tag called ``ipxe``, in order
for the DHCP server to tell the client to get the boot.ipxe script via
HTTP. Otherwise, if the tag isn't there, the DHCP server will tell the
DHCP client to chainload the iPXE image (undionly.kpxe).
The Networking service needs to be configured to create this DHCP tag,
since it isn't created by default.
#. Create a custom ``dnsmasq.conf`` file with a setting for the ipxe tag. For
example, create the file ``/etc/dnsmasq-ironic.conf`` with the content::
# Create the "ipxe" tag if request comes from iPXE user class
dhcp-userclass=set:ipxe,iPXE
# Alternatively, create the "ipxe" tag if request comes from DHCP option 175
# dhcp-match=set:ipxe,175
#. In the Networking service DHCP Agent configuration file (typically located at
/etc/neutron/dhcp_agent.ini), set the custom ``/etc/dnsmasq-ironic.conf``
file as the dnsmasq configuration file::
[DEFAULT]
dnsmasq_config_file = /etc/dnsmasq-ironic.conf
#. Restart the ``neutron-dhcp-agent`` process::
service neutron-dhcp-agent restart
IPMI support
------------
If using the IPMITool driver, the ``ipmitool`` command must be present on the
service node(s) where ``ironic-conductor`` is running. On most distros, this
is provided as part of the ``ipmitool`` package. Source code is available at
http://ipmitool.sourceforge.net/
Note that certain distros, notably Mac OS X and SLES, install ``openipmi``
instead of ``ipmitool`` by default. THIS DRIVER IS NOT COMPATIBLE WITH
``openipmi`` AS IT RELIES ON ERROR HANDLING OPTIONS NOT PROVIDED BY THIS TOOL.
Check that you can connect to and authenticate with the IPMI
controller in your bare metal server by using ``ipmitool``::
ipmitool -I lanplus -H <ip-address> -U <username> -P <password> chassis power status
<ip-address> = The IP of the IPMI controller you want to access
*Note:*
#. This is not the bare metal node's main IP. The IPMI controller
should have its own unique IP.
#. In case the above command doesn't return the power status of the
bare metal server, check for these:
- ``ipmitool`` is installed.
- The IPMI controller on your bare metal server is turned on.
- The IPMI controller credentials passed in the command are right.
- The conductor node has a route to the IPMI controller. This can be
checked by just pinging the IPMI controller IP from the conductor
node.
.. note::
If there are slow or unresponsive BMCs in the environment, the retry_timeout
configuration option in the [ipmi] section may need to be lowered. The
default is fairly conservative, as setting this timeout too low can cause
older BMCs to crash and require a hard-reset.
Bare Metal service supports sending IPMI sensor data to Telemetry with pxe_ipmitool,
pxe_ipminative, agent_ipmitool, agent_pyghmi, agent_ilo, iscsi_ilo, pxe_ilo,
and with pxe_irmc driver starting from Kilo release. By default, support for
sending IPMI sensor data to Telemetry is disabled. If you want to enable it,
you should make the following two changes in ``ironic.conf``:
* ``notification_driver = messaging`` in the ``DEFAULT`` section
* ``send_sensor_data = true`` in the ``conductor`` section
If you want to customize the sensor types which will be sent to Telemetry,
change the ``send_sensor_data_types`` option. For example, the below
settings will send temperature, fan, voltage and these three sensor types
of data to Telemetry:
* send_sensor_data_types=Temperature,Fan,Voltage
If we use default value 'All' for all the sensor types which are supported by
Telemetry, they are:
* Temperature, Fan, Voltage, Current
Configure node web console
--------------------------
See `Configuring Web or Serial Console`_.
.. _`Configuring Web or Serial Console`: http://docs.openstack.org/developer/ironic/deploy/console.html
.. _boot_mode_support:
Boot mode support
-----------------
The following drivers support setting of boot mode (Legacy BIOS or UEFI).
* ``pxe_ipmitool``
The boot modes can be configured in Bare Metal service in the following way:
* When no boot mode setting is provided, these drivers default the boot_mode
to Legacy BIOS.
* Only one boot mode (either ``uefi`` or ``bios``) can be configured for
the node.
* If the operator wants a node to boot always in ``uefi`` mode or ``bios``
mode, then they may use ``capabilities`` parameter within ``properties``
field of an bare metal node. The operator must manually set the appropriate
boot mode on the bare metal node.
To configure a node in ``uefi`` mode, then set ``capabilities`` as below::
ironic node-update <node-uuid> add properties/capabilities='boot_mode:uefi'
Nodes having ``boot_mode`` set to ``uefi`` may be requested by adding an
``extra_spec`` to the Compute service flavor::
nova flavor-key ironic-test-3 set capabilities:boot_mode="uefi"
nova boot --flavor ironic-test-3 --image test-image instance-1
If ``capabilities`` is used in ``extra_spec`` as above, nova scheduler
(``ComputeCapabilitiesFilter``) will match only bare metal nodes which have
the ``boot_mode`` set appropriately in ``properties/capabilities``. It will
filter out rest of the nodes.
The above facility for matching in the Compute service can be used in
heterogeneous environments where there is a mix of ``uefi`` and ``bios``
machines, and operator wants to provide a choice to the user regarding
boot modes. If the flavor doesn't contain ``boot_mode`` and ``boot_mode``
is configured for bare metal nodes, then nova scheduler will consider all
nodes and user may get either ``bios`` or ``uefi`` machine.
.. _choosing_the_disk_label:
Choosing the disk label
-----------------------
.. note::
The term ``disk label`` is historically used in Ironic and was taken
from `parted <https://www.gnu.org/software/parted>`_. Apparently
everyone seems to have a different word for ``disk label`` - these
are all the same thing: disk type, partition table, partition map
and so on...
Ironic allows operators to choose which disk label they want their
bare metal node to be deployed with when Ironic is responsible for
partitioning the disk; therefore choosing the disk label does not apply
when the image being deployed is a ``whole disk image``.
There are some edge cases where someone may want to choose a specific
disk label for the images being deployed, including but not limited to:
* For machines in ``bios`` boot mode with disks larger than 2 terabytes
it's recommended to use a ``gpt`` disk label. That's because
a capacity beyond 2 terabytes is not addressable by using the
MBR partitioning type. But, although GPT claims to be backward
compatible with legacy BIOS systems `that's not always the case
<http://www.rodsbooks.com/gdisk/bios.html>`_.
* Operators may want to force the partitioning to be always MBR (even
if the machine is deployed with boot mode ``uefi``) to avoid breakage
of applications and tools running on those instances.
The disk label can be configured in two ways; when Ironic is used with
the Compute service or in standalone mode. The following bullet points
and sections will describe both methods:
* When no disk label is provided Ironic will configure it according
to the `boot mode <boot_mode_support_>`_; ``bios`` boot mode will use
``msdos`` and ``uefi`` boot mode will use ``gpt``.
* Only one disk label - either ``msdos`` or ``gpt`` - can be configured
for the node.
When used with Compute service
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
When Ironic is used with the Compute service the disk label should be
set to node's ``properties/capabilities`` field and also to the flavor
which will request such capability, for example::
ironic node-update <node-uuid> add properties/capabilities='disk_label:gpt'
As for the flavor::
nova flavor-key baremetal set capabilities:disk_label="gpt"
When used in standalone mode
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
When used without the Compute service, the disk label should be set
directly to the node's ``instance_info`` field, as below::
ironic node-update <node-uuid> add instance_info/capabilities='{"disk_label": "gpt"}'